I've had some great constructive criticism feedback from @40wattstudio following the public beta release, and realised that a lot of the issues raised are due to design decisions made way back in the history of DragonMail's development, and in some cases are left-overs from previous projects. So I thought I'd go into them a bit, perhaps as a cautionary tale of being prepared to throw things out when the scope of your project changes.

I've already covered the development history of DragonMail, so here's the TL;DR: This is the second version of DragonMail. The first was set on a chain of non-randomly generated islands (another Unity Store asset) and had atrocious performance. That in turn lifted a lot of code from an earlier game set on the same set of islands, but one at a time, with a non-flying player character. The changes in the game from that earlier DragonMail and this one were really just (a) replacing the fixed landscape with a procedurally generated one and (b) replacing six distinct islands with one crater. With that in mind, it's easier to explain some of what's going on.

Is it really necessary to have 3 flight speeds?

This comes from the island chain version, where normal (slow) was (as it is now) useful for manouevering around locations and fast for moving between locations on the island. The third speed was a "warp" speed faster than the current "superspeed" which could only be used over open water and got you between islands in a sensible amount of time. The crater version kept the three speeds, but toned down "warp speed" to "superspeed" to make it manageable in a more confined landscape. (And introduced the concept of superspeed dropping off with altitude to make exploration a bit more challenging: if you want to hurtle around, you have to do it close to the ground and avoid hitting things. "Easy" mode lifts this restriction, as well as increasing your ceiling.)

So I could very well just drop "fast", and maybe tone "super" down a bit further. Hurtling around is fun, but so is being able to take in the landscape you're flying over without going at a complete crawl. A long-term goal is to reintroduce the island chain (but with procedurally generated islands) at which point three speeds will make sense again. But for now, it does only really need two.

I never would have guessed that the arrow corresponds to time of day (or why the time of day even matters to begin with).
The N for North on the compass could use more contrast to stand out.

These are both cases of reusing (or perhaps abusing) Asset Store assets beyond their intended purposes. I wanted a time of day indicator other than the position of the sun (which is important because at night most NPCs will go to sleep, and you can't deliver to them) and thought something clock-face-like would work. So I grabbed an arrow icon from a weapons inventory (which is where a lot of the icons come from) and dropped it in. v0.5.5 replaces it with a sun icon I threw together (graphics are not my strong suit) which is hopefully clearer.

The compass comes from a drop-in minimap system, which turned out to be both confusingly unhelpful and have a huge performance cost. But I loved this frame (and the character markers in the local map are from this asset too), and thought a compass would be useful. OK, I'll admit it's difficult to interpret, but the indistinct N doesn't help. I've tweaked it in v0.5.5, but honestly I'm still not completely convinced by it.

The text crawls by really slow at the very beginning of the game. A skip feature would be nice or some way to speed up the text narrative.

This is the TextMeshPro Teletype example, which has been sat in my codebases since that earlier non-DragonMail game, with minor modifications to make it integrate better with the games. Reading speed is always a personal thing -- I can't remember how I've adjusted it compared to the original, but the current rate is slightly too slow for me and I've played commercial games which are much slower. The general idea is to match the text reveal to the speed of spoken dialogue (even if there isn't any). I think the real problem here is that once you've seen the "Welcome" crawl once it gets really tedious, so I have already put in a skip (right click or backspace). But it could be better, and also the general reveal speed could be made adjustable.

While developing the game, the blue and orange altitude bars may be of use to you, but from a player perspective I don't feel like any of that information is necessary.

I'm inclined to disagree about the orange (height above ground), but the existence of the blue (height above "sea level") is a case of "I've got this widget and I'm going to use it." In this case, it's something I created for the non-DragonMail predecessor as a health/stamina meter, with a red bar nested in a green one. When I pulled the HUD code from that game into DragonMail for the dialogue box, the meter came along with it, and as I was finding height above ground difficult to judge decided it would be a good idea to repurpose one of the bars as an altimeter. But as I had two bars, I couldn't just let the other once go, so decided to make the other one a "height above sea level" altimeter. Remember that this version of the game was set on islands, so it made a little bit more sense. Once the setting moved to a crater, I decided to have two ceilings: one to limit how far above ground you can fly to make exploration more of a challenge (ie you can't just fly high enough to get an overview of the entire crater), and one to limit your absolute altitude to stop you flying out of the crater (which can sometimes be defeated, but there's not a lot to see out there). It wasn't very clear, though (and the red/green thing was just a big strike against when I started looking at accessibility issues) so I split it into side-by-side bars (and recoloured them), skipping over the fact that I had two not because the second was particularly useful but because my original widget had two bars to make use of.

I find the Control guide a little hard to read in the circular style.

That was originally created to show the controls for use with a gamepad (which is what I'm using), and the positioning on screen corresponds (roughly) to the positions on the face of the controller. I originally intended to just swap out the graphics for the keyboard (and now keyboard+mouse) layouts, but in the end it was easier just to have a complete second (and now third) screen. So I don't have much of an excuse for not fixing this.

The correlation between names and animals isn't very obvious.

I've been trying to find a balance between making the game approachable and leaving things to be discovered. And avoid having in-game characters talk about real-world game mechanics ("Press M to look at the map"), something I always find weird. The obvious problem is that because I *wrote* the thing, it's really difficult for me to judge just how discoverable this stuff is. It was always my intention for the NPCs to have more varied dialogue, and for some of that to be hints towards, or even outright explanations of mechanics like names, how superspeed is affected by altitude, who pays the best bonuses, even why it's quite light at night and you see two shadows at dawn and dusk. And, once I'd figured out how to do it, give suggestions as to where other NPCs could be found. So I think I need to prioritise this, and maybe put some additional pointers in the "Welcome" speech.

It's always interesting to find out what other people are getting, or not getting, from the game, which is usually down to how their expectations (and sometimes hardware) align or don't with my design goals. But this set highlights that there are a lot of little things which have got into the beta release either because I'd deprioritised them, or because they're left overs from earlier iterations which made some sort of sense at the time and I'd got so used to having them around that I didn't realise that they were now just a source of confusion.

Although there's one you won't to get to see. v0.5.5 also removes the completely redundant code for handling static landscapes, much of which I've dragging around since that first game.